WO2011155068A1

WO2011155068A1 - Character generating system, character generating method, and program

Info

Publication number: WO2011155068A1
Application number: PCT/JP2010/059967
Authority: WO
Inventors: 正夫桑原; 奈生人小湊; 和満盛山
Original assignee: 株式会社アルトロン
Priority date: 2010-06-11
Filing date: 2010-06-11
Publication date: 2011-12-15
Also published as: JPWO2011155068A1; US20130120425A1; US8497869B2; JP5632469B2; EP2581881A1

Abstract

Disclosed is a character generating system, a character generating method, and a program, wherein a face of a character within a virtual space attained using a computer such as a mobile game device is easy to generate, as an animation type portrait that is based on picked-up face-image information of an object. The character generating system (10) is provided with a picked-up image information acquisition unit (14), a face-part texture generating unit (15), a shape-model generating unit (16), and a texture pasting unit (17). The face-part texture generating unit (15) has a function to generate a part texture in accordance with a face part to be pasted onto a head-shape model of a character (70) that is to become the animation type portrait that is based on a picked-up face-image information of an object, and is comprised of a feature point extraction unit (151), a part selection unit (152), a flesh-color setting unit (153), and a parts-arrangement setting unit (154).

Description

Character generation system, character generation method and program

The present invention relates to a character generation system, a character generation method, and a program for generating a character in a virtual space realized using a computer. In particular, the present invention relates to a character generation system, a character generation method, and a program for generating a character's face in a virtual space realized by using a portable game device as an animation-like portrait based on face captured image information of an object.

Conventionally, an appearance character displayed in a virtual space such as a game realized by using a computer such as a portable game device is selected by a user selecting a desired character from a plurality of characters prepared in advance. It was displayed in the virtual space as an appearance character.

Also proposed are an image processing apparatus and an image processing method for pasting a captured image taken by a user with a digital camera or the like to an appearance character displayed in a virtual space. For example, in Patent Document 1, two-dimensional image information is captured and a frame for specifying a clipping range provided with a gauge provided corresponding to the feature position of the virtual three-dimensional object is used. An image processing apparatus and an image processing method are disclosed that cut out two-dimensional image information in a predetermined range and paste the cut-out two-dimensional image information on a virtual three-dimensional object.

JP 2000-235656 A

However, as disclosed in Patent Document 1, two-dimensional image information (face-captured image information) of a face part is cut out from two-dimensional image information captured by a digital camera or the like, and a virtual three-dimensional object (head shape model) is extracted. ) Increases the amount of data for generating the character, and in a device with a small storage capacity such as a portable game device, it takes time to generate the character, or the generated character is virtually There are problems that the processing speed is slow when moving in the space, and the number of characters to appear is limited. Also, there are cases where it is not desired to display an image of the subject's face as the character's face, or to display it for personal information protection.

The present invention has been made to solve the above-described problems. The face of a character in a virtual space realized by using a computer such as a portable game device is based on the face-captured image information of the object. It is an object of the present invention to provide a character generation system, a character generation method, and a program that can be easily generated as a cartoon-like portrait.

The following invention is provided to solve the above-mentioned conventional problems.
The character generation system according to the first aspect of the present invention uses a computer to display the character face displayed on the display unit as being present in the virtual space in an animation-like manner based on the captured face image information of the object. A character generation system for generating a portrait, wherein the computer acquires the face captured image information of the object from captured image information of the object stored in a predetermined storage unit; The feature information of a predetermined facial part and the color information of a predetermined pixel are extracted from the face captured image information of the object acquired by the captured image acquisition means, and the color information of the predetermined pixel is extracted based on the extracted color information of the predetermined pixel. Similar skin color information similar to the skin color of the object that is the skin color of the head texture of the character is set, and the front of the object is set based on the extracted feature points Facial part texture generating means for selecting a similar animated part corresponding to the facial part similar to a facial part and generating a facial part texture corresponding to the facial part by setting an arrangement of the selected similar animated part And shape model generation means for generating the head shape model based on the feature points of the predetermined face part, and the head other than the face part texture pasted on the head shape model of the character Texture color information is replaced with the similar skin color information set by the face part texture generation means, and the face part texture generated by the face part texture generation means is pasted to the head shape model of the character And a sticking means.

In the character generation system according to the second aspect of the present invention, the face part texture generation means of the character generation system according to the first aspect of the present invention is the face of the object acquired by the captured image acquisition means. From the captured image information, the position information of the feature point of the predetermined face part is acquired, and the feature point extraction unit that acquires the color information of the predetermined pixel; and the feature point acquired by the feature point extraction unit Parts selection means for selecting the similar anime-like part similar to the shape of the face part of the target object from among a plurality of animation-like parts prepared in advance corresponding to the face part based on position information; Based on the color information of the predetermined pixel acquired by the feature point extraction unit, the previous color similar to the skin color of the object from among a plurality of skin color information prepared in advance Skin color setting means for selecting similar skin color information, and based on the position information of the feature points acquired by the feature point extraction means, the similar animation-like parts corresponding to the face parts selected by the parts selection means Part arrangement setting means for setting the arrangement and generating the facial part texture based on the arrangement of the set similar animation-like parts.

In the character generation system according to the third aspect of the present invention, when the part selection means of the character generation system according to the second aspect of the present invention is configured such that when the face part is composed of two basic face parts, left and right The position information of the feature points of either one of the basic face parts is converted into position information moved in line symmetry with respect to the face center line, and the average of the position information of the feature points of the two left and right basic face parts The shape of the basic face part is estimated from the value, the similar anime-like part corresponding to the estimated shape of the basic face part is selected from the plurality of anime-like parts, and the basic face having one face part is selected. When configured from parts, the shape of the face part is estimated from the position information of the feature points of the face part, and the similar anime-like parts corresponding to the estimated shape of the face part are converted into the plurality of anime-like parts. And selecting from among the over tool.

In the character generation system according to the fourth aspect of the present invention, the position where the part placement setting means of the character generation system according to the second or third aspect of the present invention is pasted on the head shape model is preset. In the facial part texture corresponding to the facial part, the nose and mouth of the similar animated part and the similar animated part of the mouth are arranged at positions in the facial part texture that are set in advance. The similar animation tone of the left and right eyes with respect to the face center line at a predetermined vertical position in the facial part texture of the eyes, with the left and right eye intervals calculated from the position information of the feature points. The vertical direction calculated from the position information of the feature point at the position in the left-right direction in the facial part texture of the eyebrow set in advance so that the part is a line target In the position, arrange the similar anime-like parts of the left and right eyebrows with respect to the center line of the face to be line targets, based on the arrangement of the similar anime-like parts corresponding to the nose, mouth, eyes, and eyebrows, The facial part texture corresponding to the nose, mouth, eyes, and eyebrows is generated.

In the character generation system according to the fifth aspect of the present invention, a position where the part placement setting means of the character generation system according to the second or third aspect of the present invention is pasted on the head shape model is preset. In the facial part texture corresponding to the facial part, the similarity in the facial part texture corresponding to the facial part based on the positional information of the feature point of the facial part acquired by the feature point extraction unit The position of the anime-like part is calculated, and the face part texture is generated based on the calculated position of the similar anime-like part.

In the character generation system according to the sixth aspect of the present invention, the part arrangement setting means of the character generation system according to the fourth or fifth aspect of the present invention includes a plurality of model face part textures corresponding to the face part. Then, a similar model face part texture similar to the arrangement of the similar anime-like parts corresponding to the face part is selected, and the selected similar model face part texture is set as the face part texture.

In a character generation system according to a seventh aspect of the present invention, the shape model generation means of the character generation system according to any one of the first to sixth aspects of the present invention is prepared in advance corresponding to a face contour. A basic head shape model similar to the face contour is selected from a plurality of contour shape models based on the feature points of the face contour, and a plurality of prepared in advance corresponding to predetermined face parts other than the face contour A similar part shape model similar to the predetermined face part is selected from the part shape models of the predetermined face part based on the feature points of the predetermined face part, and the selected basic head shape model and the similar part shape are selected. The head shape model is generated in combination with a model.

In the character generation system according to the eighth aspect of the present invention, the shape model generation means of the character generation system according to any one of the first to sixth aspects of the present invention is prepared in advance corresponding to the face contour. The head shape model similar to the face contour is selected from a plurality of reference shape models based on the feature points of the face contour.

A character generation system according to a ninth aspect of the present invention is the character generation system according to any one of the first to eighth aspects of the present invention, wherein the computer controls the display unit to perform the predetermined face. Face arrangement guide information indicating the arrangement of parts, display control means for displaying the target object on the display unit, and an imaging unit are controlled to cause the target object to be imaged, and the captured image information of the target object is the predetermined image Imaging control means to be stored in the storage unit, and the captured image acquisition means interlocks the display control means and the imaging control means to display the face arrangement guide information and the object. The face captured image information of the object obtained by imaging the object based on the face arrangement guide information while being displayed on the part is displayed in front of the object stored in the predetermined storage unit. Features that obtained from the captured image information.

A character generation system according to a tenth aspect of the present invention is the character generation system according to any one of the first to ninth aspects of the present invention, wherein the computer controls the input unit to It further includes input control means for inputting various information including captured image information as input information, and storing the captured image information of the object in the predetermined storage unit, and the captured image acquisition means includes the display control means and In conjunction with the input control means, the face arrangement guide information and the captured image information of the inputted object are displayed on the display unit, and the object based on the face arrangement guide information is displayed. Face captured image information is obtained from the captured image information of the object stored in the predetermined storage unit.

The character generation system according to an eleventh aspect of the present invention is the character generation system according to the ninth or tenth aspect of the present invention, wherein the predetermined facial part whose arrangement is indicated by the face arrangement guide information has a face outline. It contains at least.

The character generation method according to the first aspect of the present invention uses a computer to convert a character's face displayed on the display unit as being in a virtual space to an animated tone based on the face image information of the object. A character generation method for generating a portrait, wherein the computer (a) acquires the captured face image information of the object from captured image information of the object stored in a predetermined storage unit; (B) acquiring positional information of the feature points of the predetermined facial part from the face captured image information of the object acquired by the step (a) acquired by the step (a); Obtaining color information of the predetermined pixel; and (c) calculating a head shape model of the character based on the feature points of the predetermined facial part acquired in the step (a). And (d) setting similar skin color information similar to the skin color of the object to be the skin color of the head texture of the character based on the color information of the predetermined pixel acquired in the step (a). Further, a similar anime-like part corresponding to the face part similar to the face part of the target object is selected based on the feature points acquired in the step (a), and the selected similar anime-like part is selected. A step of generating a facial part texture corresponding to the facial part by setting an arrangement; and (e) pasting the facial part texture generated by the step (b) to the head shape model of the character. In addition, the color information of the skin part of the head texture pasted on the head shape model generated by the step (c) is obtained by the step (d). Characterized in that it comprises a step of replacing the constant has been the similar skin color information.

In the character generation method according to the second aspect of the present invention, the step (d) of the character generation method according to the first aspect of the present invention is characterized in that the feature points acquired by (d1) the step (b) Selecting the similar anime-like part similar to the shape of the face part of the target object from among a plurality of anime-like parts prepared in advance corresponding to the face part based on position information; (d2 ) Selecting the similar skin color information similar to the skin color of the object from a plurality of skin color information prepared in advance based on the color information of the predetermined pixel acquired in the step (b); d3) Based on the position information of the feature points acquired in the step (b), the arrangement of the similar anime-like parts corresponding to the face part selected in the step (d1) is set and set. A step of generating the face part texture based on the arrangement of the serial similar animation style part, characterized in that it comprises.

In the character generating method according to the third aspect of the present invention, when the step (d2) of the character generating method according to the second aspect of the present invention is such that the face part is composed of two basic face parts on the left and right, The position information of the feature points of one of the left and right basic face parts is converted into position information moved in line symmetry with respect to the face center line, and the position information of the feature points of the two left and right basic face parts The shape of the basic face part is estimated from an average value, the similar anime-like part corresponding to the estimated shape of the basic face part is selected from the plurality of anime-like parts, and the basic part having one face part is selected. When configured from a facial part, the shape of the facial part is estimated from position information of the feature points of the facial part, and the similar animated part corresponding to the estimated shape of the facial part is converted into the plurality of animated parts. in Characterized by al selection.

In the character generation method according to the fourth aspect of the present invention, a position where the step (d3) of the character generation method according to the second or third aspect of the present invention is pasted on the head shape model is preset. In the facial part texture corresponding to the facial part, the nose and mouth of the similar animated part and the similar animated part of the mouth are arranged at positions in the facial part texture that are set in advance. The similar animation tone of the left and right eyes with respect to the face center line at a predetermined vertical position in the facial part texture of the eyes, with the left and right eye intervals calculated from the position information of the feature points. The part is arranged to be a line target, and the position of the eyebrow in the left and right direction within the face part texture set in advance is the position in the vertical direction calculated from the position information of the feature point. Place the similar anime-like parts on the left and right eyebrows with respect to the line as the line target, and based on the arrangement of the similar anime-like parts corresponding to the nose, mouth, eyes, and eyebrows, the nose, mouth, eyes And the facial part texture corresponding to the eyebrows is generated.

In the character generation method according to the fifth aspect of the present invention, a position where the step (d3) of the character generation method according to the second or third aspect of the present invention is pasted on the head shape model is preset. In the facial part texture corresponding to the facial part, the similar animated part in the facial part texture corresponding to the facial part based on the positional information of the feature points acquired in the step (b). A position is calculated, and the facial part texture is generated based on the calculated position of the similar anime-like part.

In the character generation method according to the sixth aspect of the present invention, the step (d3) of the character generation method according to the fourth or fifth aspect of the present invention includes a plurality of model face part textures corresponding to the face part. A similar model face part texture similar to the arrangement of the similar anime-like parts corresponding to the face part is selected, and the selected similar model face part texture is set as the face part texture. And

In the character generating method according to the seventh aspect of the present invention, the step (c) of the character generating method according to any one of the first to sixth aspects of the present invention is prepared in advance corresponding to the face contour. A basic head shape model similar to the face contour is selected from a plurality of contour shape models based on the feature points of the face contour, and a plurality of prepared in advance corresponding to predetermined face parts other than the face contour A similar part shape model similar to the predetermined face part is selected from the part shape models of the predetermined face part based on the feature points of the predetermined face part, and the selected basic head shape model and the similar part shape are selected. The head shape model is generated by combining with a model.

In the character generation method according to the eighth aspect of the present invention, the step (c) of the character generation method according to any one of the first to sixth aspects of the present invention is prepared in advance corresponding to the face contour. The head shape model similar to the face contour is selected from a plurality of reference shape models based on the feature points of the face contour.

A character generation method according to a ninth aspect of the present invention is the character generation method according to any one of the first to eighth aspects of the present invention, wherein the computer shows (f) a predetermined arrangement of the face parts. While displaying the face arrangement guide information and the object on the display unit, the target object is imaged based on the face arrangement guide information, and the captured image information of the object is stored in the predetermined storage unit. A step is further provided before (a), and the step (a) stores the face captured image information of the object in the predetermined storage unit based on the face arrangement guide information. It is acquired from the captured image information of the object.

A character generation method according to a tenth aspect of the present invention is the character generation method according to any one of the first to ninth aspects of the present invention, wherein the computer (g) captures the captured image information of the object. The step of inputting and storing in the predetermined storage unit is further provided before (a), wherein the step (a) includes the face arrangement guide information indicating the arrangement of the predetermined facial part, and the step Based on the face arrangement guide information, the face captured image information of the object is stored in a predetermined storage unit while displaying the captured image information of the object input in (g) on the display unit. It is acquired from the captured image information of the stored object.

A character generation method according to an eleventh aspect of the present invention is the character generation method according to the ninth or tenth aspect of the present invention, wherein the predetermined facial part whose arrangement is indicated by the face arrangement guide information has a face outline. It contains at least.

The program according to the first aspect of the present invention performs processing for generating a character's face displayed on the display unit as existence in the virtual space as an animation-like portrait based on the captured face image information of the object. A program to be executed by a computer, characterized by causing the computer to execute processing for realizing each means of the character generation system according to any one of the first to eleventh aspects of the present invention.

According to the present invention, a character's face in a virtual space realized by using a computer such as a portable game device can be easily converted into an animation-like caricature based on face image information of an object photographed by a digital camera or the like. Can be generated. In addition, by generating the face of a character appearing in a virtual space such as a game as an animation-like portrait based on the user's face captured image information, the user can be more emotionally transferred to the character. More highly entertaining game software can be constructed. In addition, the amount of character data that uses anime-like caricatures is small compared to the amount of character data that uses photographic images, game software that speeds up the processing of games, etc., and allows more characters to appear Can be built.

2A and 2B show an example of a character 70 generated by the character generation system 10 according to an embodiment of the present invention, where FIG. 1A shows an example of a female character 70a and FIG. 2B shows an example of a male character 70b. . It is a figure which shows schematic structure of the portable game device 20 which performs the character production | generation system 10 concerning one Embodiment of this invention. It is an example of the display screen on which face arrangement guide information is displayed. It is a figure which shows an example of the system configuration | structure in the character production | generation system 10 concerning one Embodiment of this invention. It is the top view seen from the front of the head shape model 60 for demonstrating a facial component texture. (A) is a figure for demonstrating a feature point, (b) is an enlarged view of the right eye area | region 41. FIG. (A) thru | or (c) is the figure which showed the some example of the animation-like part of a female eyebrow. (A) thru | or (c) are the figures which showed the some example of the animation-like parts of a female eye. (A) thru | or (c) are the figures which showed the some example of the animation-like parts of a woman's mouth. (A) thru | or (c) is the figure which showed the some example of the animation-like parts of a male eyebrow. (A) thru | or (c) are the figures which showed the some example of the animation-like parts of a male eye. (A) thru | or (c) is the figure which showed the some example of the animation-like parts of a male mouth. It is the top view seen from the front of head shape model 60 for explaining arrangement of similar animation tone parts in face part texture 51. It is a figure for demonstrating the basic head shape model 71 and the similar part shape model 72 of a state without a nose, (a) is a top view of a face front, (b) is a top view of the face side. . It is a figure for demonstrating the head shape model 60, (a) is a top view of the face front, (b) is a top view of the face side. It is an example of the flowchart figure which shows the process sequence of the program which makes a computer perform each process of the character production | generation method concerning one Embodiment of this invention.

An embodiment of the present invention will be described with reference to the drawings. In addition, embodiment described below is for description and does not limit the scope of the present invention. Accordingly, those skilled in the art can employ embodiments in which each or all of these elements are replaced by equivalents thereof, and these embodiments are also included in the scope of the present invention.

A character generation system 10 according to an embodiment of the present invention targets a character face displayed on a display unit as a presence in a virtual space realized by software such as a game running on a computer. This is a character generation system that generates an animation-like portrait based on captured face information of an object (for example, a user). FIG. 1 shows an example of a character 70 in a virtual space realized by game software that is generated by a character generation system 10 according to an embodiment of the present invention and operates on a portable game device. ) Is an example showing a female character 70a, and (b) is an example showing a male character 70b.

Here, the face captured image information of the target is captured image information of the face portion in the captured image information of a person captured using a digital camera or the like, and an imaging unit connected to or built in the computer is used. The captured image information acquired directly from the imaging unit may be used, or the captured image information captured in the past input from the input unit connected to the computer may be used. Also, the image information may be image information obtained by partially correcting one piece of captured image information or combining a plurality of captured images into one piece of face captured image information. An animation-like caricature based on the captured face image information of the object is similar to the captured face image information of the object, and is a face like a person appearing in a cartoon or animation that emphasizes cuteness. That is.

First, a schematic configuration of a computer that executes a character generation system 10 according to an embodiment of the present invention will be described. In the present embodiment, the portable game device 20 will be described as an example of a computer.

FIG. 2 is a diagram showing a schematic configuration of the portable game device 20 that executes the character generation system 10 according to the embodiment of the present invention. As shown in FIG. 2, the portable game device 20 includes a CPU (central processing unit) 21, a ROM 22, a RAM 23, a display unit 24, an imaging unit 25, and an input unit 26.

The CPU 21 implements the character generation system 10 by reading out and executing necessary information from the ROM 22 storing software and data for realizing the character generation system 10 to be executed by the portable game device 20. The RAM 23 also functions as a data storage device and software execution work area necessary for realizing the character generation system 10 executed by the portable game device 20.

Further, the display unit 24 outputs display information (for example, information for prompting the user to perform work, face arrangement guide information, etc.) in accordance with a command from the CPU 21. Here, the face placement guide information is information for setting a position reference and a range for extracting face captured image information as a face portion in captured image information of a subject, and is display information indicating an arrangement of a predetermined facial part. It is. The predetermined face parts are parts such as eyes, nose, mouth, face outline, and eyebrows constituting the face, and include at least the face outline. FIG. 3 is an example of a display screen on which face arrangement guide information is displayed. As shown in FIG. 3, the face arrangement guide information indicates a nose position 31 and a face outline 32. The face outline 32 corresponds to the head shape model of the character 70.

Further, the imaging unit 25 images a subject in accordance with a command from the CPU 21 and stores captured image information of the captured subject in the RAM 23. The input unit 26 is, for example, an operation button, a communication device, an external storage device, or the like, and inputs input information (for example, information on the operation button operated by the user, captured image information, etc.) according to a command from the CPU 21. And stored in the RAM 23.

Next, a system configuration in the character generation system 10 according to an embodiment of the present invention will be described. FIG. 4 is a diagram showing an example of a system configuration in the character generation system 10 according to the embodiment of the present invention.

As shown in FIG. 4, the character generation system 10 includes a display control unit 11, an imaging control unit 12, an input control unit 13, a captured image information acquisition unit 14, a facial part texture generation unit 15, a shape model generation unit 16, and a texture pasting. Unit 17, character information storage unit 18, and guide information storage unit 19. The face part texture generation unit 15 includes a feature point extraction unit 151, a part selection unit 152, a skin color setting unit 153, and a part arrangement setting unit 154.

The display control unit 11 controls the display unit 24 to take out necessary information from the guide information storage unit 19 such as face placement guide information indicating the placement of a predetermined facial part, information for prompting the user to work, Display information is generated and displayed from the extracted information. The display control unit 11 controls all processes of the display unit 24, not limited to the above-described control. Further, the imaging control unit 12 controls the imaging unit 25 to image a subject, and causes the character information storage unit 18 to store captured image information of the captured subject. Further, the input control unit 13 controls the corresponding input unit 26 to input various input information (for example, captured image information captured in the past), and causes the character information storage unit 18 to store necessary information. Here, it is described that the captured image information of the subject is stored in the character information storage unit 18, but it may be stored in a temporary storage unit.

The captured image information acquisition unit 14 causes the display control unit 11 and the imaging control unit 12 to work together to display face arrangement guide information indicating the arrangement of a predetermined facial part and an object (subject) on the display unit 24. Then, the face of the subject is imaged so as to match the arrangement of the face parts in the face arrangement guide information (see FIG. 3). In particular, the face of the subject is imaged by matching the width and chin contour (lower contour) of the face contour 32 with the width and chin contour of the face of the subject.

Also, the captured image information acquisition unit 14 acquires facial captured image information from captured image information of the captured subject in order to generate an animation-like caricature based on the captured facial image information. The captured face image information is acquired from the captured image information of the subject imaged by the imaging unit 25, but may be acquired from the captured image information of the subject input from the input unit 26. In other words, the captured image information acquisition unit 14 links the display control unit 11 and the input control unit 13 to display the face arrangement guide information indicating the arrangement of predetermined facial parts and the captured image information of the subject on the display unit 24. While being displayed, the captured face image information matching the face placement guide information is acquired while relatively aligning the face part placement in the face placement guide information and the face part placement in the captured image information of the subject. .

The facial part texture generation unit 15 pastes the head shape model 60 (see FIGS. 15A and 15B) of the character 70 to be an animation-like caricature based on the face captured image information of the target object, A function for generating facial part textures corresponding to facial parts such as mouth, eyes, and eyebrows, and includes a feature point extraction unit 151, a part selection unit 152, a skin color setting unit 153, and a part arrangement setting unit 154. FIG. 5 is a plan view of the head shape model 60 for explaining the facial part texture as viewed from the front.

As shown in FIG. 5, the face part texture 51 (for example, mouth face part texture 51 a, nose face part texture 51 b, eye face part texture 51 c, eyebrow face part texture 51 d) is added to the head shape model 60. A position to be pasted is set in advance. The face part texture 51 is a texture in which a similar anime-like part 52 similar to the face part of the target object is arranged. For example, as shown in FIG. 5, the eye face part texture 51c is a texture in which two similar anime-like parts 52 selected based on the basic face parts of the right eye and the left eye are arranged. In FIG. 5, the facial part textures 51 are arranged so as not to overlap each other, but the facial part textures 51 may be arranged so as to overlap if at least different similar anime-like parts 52 do not overlap each other. .

In this specification, when there are two left and right eyes like the eyes, the right eye and the left eye are called basic face parts, and the left and right eyes are collectively called face parts. In addition, when there is only one like a mouth, the mouth is called a basic face part and / or a face part. The texture pasted on the head shape model 60 is referred to as a head texture 50, and the head texture 50 other than the face part texture 51 is referred to as a blank portion 54. That is, the head texture 50 pasted on the head shape model 60 is composed of the face part texture 51 and the margin part 54.

The feature point extraction unit 151 of the face part texture generation unit 15 extracts feature points of predetermined facial parts and color information of predetermined pixels from the face captured image information acquired by the captured image information acquisition unit 14. That is, the position information of the feature points of a predetermined facial part is acquired from the captured face image information acquired by the captured image information acquisition unit 14, and the color information of the predetermined pixel is acquired. FIG. 6A is a diagram for explaining feature points, and FIG. 6B is an enlarged view of the right eye region 41. In FIGS. 6A and 6B, eyebrows, eyes, nose, mouth, and facial contours are set as predetermined facial parts. The points indicated by black circles indicate feature points. The feature points are points that characterize the shape of a predetermined facial part. For example, the seven black circles in the right eye region 41 are the right eye feature points 43a to 43g. The color information of the predetermined pixel is, for example, the color information of the pixel at the nose vertex 42 and the color information of the pixels at the

positions

43a and 44a under both eyes. The right and left of the right eye and the left eye are the right side and the left side of the drawing. The color information is color information represented by RGB, YUV, or the like.

The part selection unit 152 of the facial part texture generation unit 15 estimates the shape of the facial part based on the position information of the feature point of the predetermined facial part extracted by the feature point extraction unit 151, and the estimated facial part A similar anime-like part 52 similar to the shape of is selected from a plurality of animation-like parts prepared in advance. Here, the anime-like parts are pictures that deform the corresponding face parts. A plurality of anime-like parts are prepared based on gender, race, age, and the like. 7 to 9 are diagrams showing examples of anime-like parts corresponding to facial parts. FIGS. 7A to 7C are diagrams showing a plurality of examples of anime-like parts for female eyebrows. 8A to 8C are diagrams showing a plurality of examples of anime-like parts for female eyes. FIGS. 9A to 9C are diagrams showing a plurality of examples of anime-like parts of a female mouth. FIGS. 10A to 10C are diagrams showing a plurality of examples of anime-like parts for male eyebrows. FIGS. 11A to 11C are diagrams showing a plurality of examples of anime-like parts of male eyes. 12 (a) to 12 (c) are diagrams showing a plurality of examples of anime-like parts of a male mouth.

As shown in FIGS. 6A and 6B, for example, in the case of an eye, the width of the eye (the distance between the points 43b and 43f), the height of the eye (the distance between the

points

43a and 43d), and the eyelet The shape of the eye according to the inclination of the eye (such as a line segment 49 connecting the points 43g and 43b and an angle θ between the line segment 48 extending horizontally from the point 43g in the left-right direction (X direction)). Is estimated. Therefore, the anime-like part closest to the estimated eye width, eye height, and eye inclination among the anime-like parts of the plurality of eyes becomes the similar anime-like part 52 of the eyes.

In addition, since the eyebrows and eyes are composed of two basic face parts on the left and right, if the shape is estimated separately on the left and right, the left and right may have different shapes. However, when considering the balance of the entire face, if the left and right sides have the same shape (a shape that is line-symmetric with respect to the face center line), the balance of the entire face will not be lost. The position information is converted into position information that is moved symmetrically with respect to the face center line, that is, the position of the eye is calculated from the average of the position information of the feature points of the two right eyes after being horizontally reversed and converted into the position of the right eye. The position information of the points is calculated, and the shape of the eyes is estimated. The same applies to the eyebrows. For the nose, mouth and face outlines, the shape is estimated from the extracted feature point position information. In the case of eyes, the color information of the pupil may be extracted by the feature point extraction unit 151, and the color information of the pupil of the similar anime-like part 52 may be changed based on the extracted color information.

The skin color setting unit 153 of the face part texture generation unit 15 uses color information similar to the skin color of the target object (hereinafter referred to as similar skin color information) based on the color information of the predetermined pixel extracted by the feature point extraction unit 151. The calculated similar skin color information is set as the color information of the head texture 50 of the character 70. For example, when the color information of the predetermined pixel extracted by the feature point extraction unit 151 is the color information of the pixel of the nose vertex 42 and the color information of the

pixels

43a and 44a under both eyes, the colors of the three pixels An average of the information is calculated, and the calculated color information is set as similar skin color information.

The part arrangement setting unit 154 of the facial part texture generation unit 15 selects the part selection unit 152 in the facial part texture 51 based on the position information of the feature points of the predetermined facial part extracted by the feature point extraction unit 151. The arrangement of the similar anime-like part 52 is set, and the face part texture 51 is generated based on the set position of the similar anime-like part 52.

FIG. 13 is a plan view seen from the front of the head shape model 60 for explaining the arrangement of the similar animation-like parts 52 in the face part texture 51. FIG. As shown in FIG. 13, in each of the

facial part textures

51 a, 51 b, 51 c, 51 d to be pasted on the head shape model 60, the point 61 is the arrangement of the nose-like animation part 52, and the point 62 is the mouth similarity The arrangement of the anime-like parts 52, the point 63 the arrangement of the similar-animation-like parts 52 of the right eye, the point 64 the arrangement of the similar-animation-like parts 52 of the left-eye, and the point 65 the arrangement of the similar-animation-like parts 52 of the right eyebrow. A point 66 indicates the arrangement of the similar anime-like part 52 of the left eyebrow.

The position of the head shape model 60 to which the

facial part textures

51 a, 51 b, 51 c, 51 d are pasted is set in advance corresponding to the head shape model 60.
With respect to the nose face part texture 51b and the mouth face part texture 51b, a point 61 that is the arrangement of the nose-like animation-like part 52 and a point 62 that is the arrangement of the mouth-like animation-like part 52 are preset. Yes. As for the arrangement of the eye-like anime-like part 52 with respect to the eye-face part texture 51c, the position of the eye-like anime-like part 52 in the vertical direction (height direction: Y direction) is set in advance, and the left-right direction (width) The position of (direction: X direction) is calculated from the feature points of both eyes. That is, the distance 67 between the similar anime-like parts 52 of both eyes is calculated from the feature points of both eyes at the preset vertical position (height direction: Y direction) of the similar anime-like parts 52 of both eyes. The similar anime-like parts 52 are arranged so as to be line targets with respect to the face center line 69. In addition, regarding the arrangement of the similar animation-like part 52 of the eyebrow with respect to the facial part texture 51d of the eyebrow, the horizontal position of the similar animation-like part 52 of the eyebrow is preset, and the vertical position is calculated from the feature points of both eyebrows. To do. The vertical position of the similar animation-like parts 52 of both eyebrows is calculated from the feature points of both eyebrows, and the similar animation-like parts 52 of both eyebrows are arranged so as to be a line target with respect to the face center line 69.

The arrangement of the similar anime-like parts 52 with respect to the face part texture 51 is not limited to the arrangement process described above, and the arrangement of the similar anime-like parts 52 in all the face part textures 51 is determined based on the feature points of the face parts. It may be calculated. Further, the arrangement of the similar animation-like parts 52 with respect to the face part texture 51 may be fixed in advance.

In addition to the facial part texture 51 in which the similar animated part 52 is arranged in the facial part texture 51 based on the arrangement of the similar animated part 52 with respect to the facial part texture 51, a plurality of model facial part textures corresponding to the facial part are included. The similar model face part texture similar to the arrangement of the similar anime-like part 52 with respect to the face part texture 51 may be selected, and the selected similar model face part texture may be set as the face part texture 51.

As described above, the facial part texture generation unit 15 generates the facial part texture 51 corresponding to the facial part of the cartoon-like portrait based on the captured face image information of the subject. Then, the generated facial part texture 51 is stored in the character information storage unit 18.

The shape model generation unit 16 has a function of generating a head shape model 60 based on feature points of predetermined face parts. Further, the margin part 54 of the head texture 50 is pasted on the generated head shape model 60.

The head contour component shape similar to the face contour based on the facial contour feature points acquired by the facial component texture generation unit 15 from the plurality of contour shape models stored in advance in the head shape model 60 A model (hereinafter referred to as a basic head shape model) 71 is selected, and similar to the face part among a plurality of part shape models corresponding to a predetermined face part stored in the head shape model 60 in advance. A part shape model (hereinafter referred to as a similar part shape model) 72 of the face part is selected. Then, the head shape model 60 is generated by combining (combining) the selected basic head shape model 71 and the similar shape model 72. Here, the basic head shape model 71 is a head shape model in a state where there is no part shape model portion corresponding to a predetermined face part. That is, for example, when the predetermined facial part is the nose, the head shape model without the nose is the basic head shape model 71.

14A and 14B are diagrams for explaining the basic head shape model 71 and the similar part shape model 72 in a state where no nose is present. FIG. 14A is a plan view of the face front, and FIG. A plan view is shown. FIGS. 15A and 15B are diagrams for explaining the head shape model 60. FIG. 15A is a plan view of the face front, and FIG. 15B is a plan view of the face side. 14A, 14 </ b> B, 15 </ b> A, and 15 </ b> B, the nose similar part shape model 72 will be described as an example. 14A and 14B, a combination of the basic head shape model 71 selected based on the feature points of the face contour and the similar part shape model 72 of the nose selected based on the feature points of the nose Thus, the head shape model 60 shown in FIGS. 15A and 15B is generated.

Although the generation of the head shape model 60 described above has been described with respect to the case where the head shape model 60 is generated in combination with the basic head shape model 71 and the similar part shape model 72, the head shape model 60 is stored in advance in the head shape model 60. A head shape model 60 similar to the face contour may be selected from a plurality of reference shape models based on the feature points of the face contour acquired by the face part texture generation unit 15. Further, one of the head shape models 60 set in advance may be stored in the character information storage unit 18 and the head shape model 60 may be taken out from the character information storage unit 18.

The texture pasting unit 17 pastes the face part texture 51 generated by the face part texture generation unit 15 on the head shape model 60 generated by the shape model generation unit 16 and the margin part 54 of the head texture 50. The color information and the color information of the skin part of the face part texture 51 are replaced with similar skin color information set by the skin color setting unit 153 of the face part texture generation unit 15.

Note that a plurality of parts corresponding to hairstyles, clothes, accessories, etc. are stored in advance in the character information storage unit 18 in accordance with the situation, sex, etc. of the game, etc. You can also set parts. In the texture pasting unit 17 described above, the color information of the marginal part 54 of the head texture 50 and the color information of the skin part of the face part texture 51 are replaced with similar skin color information. The shape model generation unit 16 performs a process of replacing the color information of the margin part 54 with the similar skin color information, and the process of replacing the color information of the skin part of the face part texture 51 with the similar skin color information. The arrangement setting unit 154 may perform this.

Next, the start timing of the character generation system 10 according to an embodiment of the present invention will be described. For example, in the case of a role playing game (RPG), the activation timing of the character generation system 10 according to the embodiment of the present invention is started by starting the character generation system 10 before starting the RPG. In addition to generating the faces of all the characters 70 that have previously appeared in the virtual space of the RPG, the character generation system 10 is used when a new character 70 appears in the virtual space during the progress of the RPG game. The face of the appearing character 70 can also be generated by operating it. That is, the character generation system 10 can be activated at a user's desired activation timing.

Face-captured image information of an object obtained by photographing the face of the character 70 in the virtual space realized by using the portable game device 20 with a digital camera or the like by the character generation system 10 according to the embodiment of the present invention described above. It can be easily generated as an anime-like caricature based on. Further, by generating the face of the character 70 appearing in a virtual space such as a game as an animation-like caricature based on the user's captured face image information, the user can more empathize with the character 70. It is possible to build game software with higher entertainment. Further, the data amount of the character 70 using the cartoon-like caricature is smaller than the data amount of the character 70 using the photographic image, so that the processing speed of the game or the like is increased, or more characters 70 appear. Game software can be built.

Next, a character generation method according to an embodiment of the present invention will be described.
FIG. 16 is an example of a flowchart showing a processing procedure of a program that causes a computer to execute each step of the character generation method according to the embodiment of the present invention.

To generate a character, first, face-captured image information that is a target for generating an animation-like portrait is acquired (step 101: S101). Here, while the face arrangement guide information indicating the arrangement of a predetermined face part and the subject are displayed on the display unit 24, the face of the subject is imaged by the imaging unit 25 so as to match the arrangement of the face parts of the face arrangement guide information. The captured image information of the captured subject is acquired. Furthermore, face captured image information is acquired from the captured image information of the captured subject. Here, captured image information of the subject imaged by the imaging unit 25 is used, but captured image information input from the input unit 26 may be used. Also, the image information may be image information obtained by partially correcting one piece of captured image information or combining a plurality of captured images into one piece of face captured image information.

Next, feature points of predetermined face parts and color information of predetermined pixels are extracted from the captured face image information (step 102: S102). That is, the position information of feature points of a predetermined facial part is acquired from the captured face image information acquired in the step 101, and the color information of the predetermined pixel is acquired. Here, for example, the predetermined facial parts are eyebrows, eyes, nose, mouth, and face outline. Further, for example, the color information of the predetermined pixel is the color information of the pixel at the nose vertex 42 in FIGS. 6A and 6B and the color information of the

pixels

43a and 44a under both eyes.

Next, the head shape model 60 is generated based on the position information of the feature points of the predetermined facial part extracted in the step 102 (step 103: S103). As described in the shape model generation 16 above, for example, the facial contour feature points acquired by the facial part texture generation unit 15 from the plurality of contour shape models stored in advance in the head shape model 60 are used. Based on this, a basic head shape model 71 similar to the face contour is selected, and similar to the face part from among a plurality of part shape models corresponding to predetermined face parts stored in the head shape model 60 in advance. The selected similar part shape model 72 is selected, and the head shape model 60 is generated by combining the selected basic head shape model 71 and the similar shape model 72.

Further, the head shape model similar to the face contour based on the facial contour feature points acquired by the face part texture generation unit 15 from the plurality of reference shape models stored in advance in the head shape model 60. 60 may be selected. Further, one head shape model 60 may be stored in the character information storage unit 18 in advance, and the head shape model 60 may be taken out from the character information storage unit 18.

Next, the shape of the face part is estimated based on the position information of the feature points of the predetermined face part extracted in step 102 (step 104: S104). As described above with reference to FIGS. 6A and 6B, for example, in the case of eyes, the eye width (the distance between the points 43b and 43f) and the eye height (the distance between the

points

43a and 43d). The shape of the eye is estimated based on the inclination of the eye indicating the leaning, fishing, etc. (angle θ between the line connecting point 43g and point 43b and the line extending horizontally from point 43g in the horizontal direction). .

Next, a similar anime-like part 52 similar to the shape of the face part estimated in the step 104 is selected from a plurality of animation-like parts prepared in advance (step 105: S105). Next, similar skin color information similar to the skin color of the object is calculated based on the color information of the predetermined pixel extracted in step 102 (step 106: S106). As described above with reference to FIGS. 6A and 6B, for example, the color information of the extracted predetermined pixel is the color information of the pixel of the nose vertex 42, the

pixels

43a and 44a below both eyes. When it is color information, the average of the color information of three pixels is calculated, and the calculated color information is set as similar skin color information.

Next, based on the position information of the feature points of the predetermined facial part extracted in the process of step 102, the arrangement of the similar animated parts 52 in the facial part texture 51 is set, and the similar animated parts 52 of the set similar animated parts 52 are set. The face part texture 51 is generated based on the position (step 107: S107). As described above with reference to FIG. 13, with respect to the nose facial part texture 51b and the mouth facial part texture 51b, the point 61, which is the arrangement of the nose-like animation part 52, and the arrangement of the mouth-like animation part 52 The point 62 is set in advance. As for the arrangement of the eye-like anime-like part 52 with respect to the eye face part texture 51c, the vertical position of the eye-like anime-like part 52 is preset, and the left-right direction position is calculated from the feature points of both eyes. . The distance 67 between the similar anime-like parts 52 of both eyes is calculated from the feature points of both eyes, and the similar anime-like parts 52 of both eyes are arranged so as to be a line target with respect to the face center line 69. In addition, regarding the arrangement of the similar animation-like part 52 of the eyebrow with respect to the facial part texture 51d of the eyebrow, the horizontal position of the similar animation-like part 52 of the eyebrow is preset, and the vertical position is calculated from the feature points of both eyebrows. To do. The vertical position of the similar animation-like parts 52 of both eyebrows is calculated from the feature points of both eyebrows, and the similar animation-like parts 52 of both eyebrows are arranged so as to be a line target with respect to the face center line 69. Further, the arrangement processing is not limited to the above-described arrangement processing, and the arrangement of the similar animated parts 52 in all the facial part textures 51 may be calculated based on the feature points of the facial parts. Further, the arrangement of the similar animation-like parts 52 with respect to the face part texture 51 may be fixed in advance.

Further, based on the arrangement of the similar anime-like parts 52 with respect to the face-part texture 51, not only the face-part texture 51 in which the similar-animation-like parts 52 are arranged in the face-part texture 51, but also a plurality of model face parts corresponding to the face parts. A similar model face part texture similar to the arrangement of the similar anime-like part 52 with respect to the face part texture 51 may be selected from the textures, and the selected similar model face part texture may be set as the face part texture 51.

Finally, the facial part texture 51 generated by the step 107 is pasted on the head shape model 60 generated by the step 103, and the color information and the facial part texture of the margin part 54 of the head texture 50 are pasted. The color information of the skin portion 51 is replaced with the similar skin color information set in the step 106 (step 108: S108). As a result, the character 70 with the animation-like portrait attached is generated. In step 108, the color information of the margin part 54 of the head texture 50 and the color information of the skin part of the face part texture 51 are replaced with similar skin color information. The process of replacing the color information of the part 54 with the similar skin color information may be performed in the process of step 103, and the process of replacing the color information of the skin part of the facial part texture 51 with the similar skin color information may be performed in the process of step 107.

By using the character generation method according to the embodiment of the present invention described above, the face of the character 70 in the virtual space realized by using the portable game device 20 is converted into face-captured image information of an object photographed by a digital camera or the like. It can be easily generated as a cartoon-like caricature based on it. Further, by generating the face of the character 70 appearing in a virtual space such as a game as an animation-like caricature based on the user's captured face image information, the user can more empathize with the character 70. It is possible to build game software with higher entertainment. Further, the data amount of the character 70 using the cartoon-like caricature is smaller than the data amount of the character 70 using the photographic image, and the processing speed of the game or the like is increased, or more characters 70 appear. Game software can be built.

The character generation system 10 according to the embodiment of the present invention described above is a system using the portable game device 20, but is not limited to this, and is not limited to this, and is a game machine for arcade (arcade game machine), a home game. Applicable to the generation of 2D or 3D characters represented as existence in a virtual space realized by using a computer, a mobile phone, a stand-alone computer, a workstation computer system, a network computer system, etc. it can.

DESCRIPTION OF SYMBOLS 10: Character generation system 11: Display control part 12: Imaging control part 13: Input control part 14: Captured image information acquisition part 15: Facial part texture generation part 16: Shape model generation part 17: Texture sticking part 18: Character information storage Unit 19: Guide information storage unit 50: Head texture 51: Face part texture 52: Similar animation part 54: Margin part 60: Head shape model 70: Character 71: Basic head shape model 72: Similar part shape model 151 : Feature point extraction unit 152: Parts selection unit 153: Skin color setting unit 154: Parts placement setting unit

Claims

A character generation system that uses a computer to generate a face of a character displayed on a display unit as a presence in a virtual space as an animation-like caricature based on face captured image information of an object,
The computer is
Captured image acquisition means for acquiring the captured face image information of the object from the captured image information of the object stored in a predetermined storage unit;
Feature information of a predetermined facial part and color information of a predetermined pixel are extracted from the face captured image information of the object acquired by the captured image acquisition means, and the character is based on the extracted color information of the predetermined pixel. Similar skin color information similar to the skin color of the object that is the skin color of the head texture is set, and the similarity corresponding to the facial part similar to the facial part of the target object based on the extracted feature points A facial part texture generating means for selecting a cartoon part and generating a facial part texture corresponding to the facial part by setting the arrangement of the selected similar animated part;
Shape model generation means for generating the head shape model based on the feature points of a predetermined face part;
The color information of the head texture other than the face part texture pasted on the head shape model of the character is replaced with the similar skin color information set by the face part texture generation means, and the face part texture generation means Texture pasting means for pasting the generated facial part texture to the head shape model of the character;
A character generation system comprising:
The facial part texture generation means includes:
Feature point extraction means for acquiring position information of the feature point of the predetermined facial part and color information of the predetermined pixel from the face captured image information of the object acquired by the captured image acquisition means When,
Based on the position information of the feature points acquired by the feature point extraction unit, the shape is similar to the shape of the face part of the target object from among a plurality of animation-like parts prepared in advance corresponding to the face part Parts selection means for selecting the similar anime-like parts;
Skin color setting means for selecting the similar skin color information similar to the skin color of the target object from a plurality of skin color information prepared in advance based on the color information of the predetermined pixel acquired by the feature point extraction means;
Based on the position information of the feature points acquired by the feature point extraction means, the arrangement of the similar animation-like parts corresponding to the face parts selected by the parts selection means is set, and the set similar animation style Parts placement setting means for generating the facial part texture based on the placement of the parts;
The character generation system according to claim 1, further comprising:
The parts selection means is
When the face part is composed of two left and right basic face parts, the position information of the feature point of either the left or right basic face part is converted into position information moved in line symmetry with respect to the face center line. The shape of the basic face part is estimated from the average value of the position information of the feature points of the two left and right basic face parts, and the similar anime-like parts corresponding to the estimated shape of the basic face part are determined as the plurality of animations. Select from the key parts,
When the face part is composed of one basic face part, the shape of the face part is estimated from position information of the feature points of the face part, and the similar animation tone corresponding to the estimated shape of the face part The character generation system according to claim 2, wherein a part is selected from the plurality of anime-like parts.
The parts arrangement setting means
In the facial part texture corresponding to the facial part, the position pasted on the head shape model is preset,
The nose and mouth similar animation-like parts and the mouth similar animation-like parts are arranged in positions in the face part texture of the nose and mouth,
The similar anime-like parts of the left and right eyes with respect to the face center line at the vertical position in the face part texture set in advance with the distance between the left and right eyes calculated from the position information of the feature points Is placed to be a line target,
The similar animated parts of the left and right eyebrows with respect to the face center line at the position in the vertical direction calculated from the position information of the feature point at the position in the left and right direction in the face part texture of the eyebrow set in advance. Arrange to be a line target,
4. The face part texture corresponding to the nose, mouth, eyes, and eyebrows is generated based on the arrangement of the similar anime-like parts corresponding to the nose, mouth, eyes, and eyebrows. The character generation system described in 1.
The parts arrangement setting means
In the facial part texture corresponding to the facial part, the position pasted on the head shape model is preset,
Calculating the position of the similar anime-like part in the face part texture corresponding to the face part based on the position information of the feature point of the face part acquired by the feature point extracting means;
The character generation system according to claim 2, wherein the facial part texture is generated based on the calculated position of the similar anime-like part.
The parts arrangement setting means
A similar model face part texture similar to the arrangement of the similar animation-like parts corresponding to the face part is selected from a plurality of model face part textures corresponding to the face part, and the selected similar model face part texture is The character generation system according to claim 4, wherein the character generation system is set as a facial part texture.
The shape model generating means is
A basic head shape model similar to the face contour is selected from a plurality of contour shape models prepared in advance corresponding to the face contour, based on the feature points of the face contour,
A similar part shape model similar to the predetermined face part based on the feature points of the predetermined face part from among a plurality of part shape models prepared in advance corresponding to the predetermined face part other than the face contour Select
The character generation system according to claim 1, wherein the head shape model is generated by combining the selected basic head shape model and the similar part shape model.
The shape model generating means is
2. The head shape model similar to a face contour is selected based on the feature points of the face contour from a plurality of reference shape models prepared in advance corresponding to the face contour. The character generation system of any one of thru | or 6.
The computer is
Display control means for controlling the display unit to display face arrangement guide information indicating the arrangement of the predetermined facial parts and the object on the display unit;
Imaging control means for controlling the imaging unit to image the object and storing the captured image information of the object in the predetermined storage unit;
Further comprising
The captured image acquisition means links the display control means and the imaging control means to display the face placement guide information and the object on the display unit, based on the face placement guide information. 9. The face picked-up image information of the object picked up by the object is acquired from the picked-up image information of the object stored in the predetermined storage unit. The character generation system according to item 1.
The computer is
An input control unit that controls the input unit to input various information including the captured image information of the object as input information, and stores the captured image information of the object in the predetermined storage unit;
The captured image acquisition means links the display control means and the input control means to display the face arrangement guide information and the input captured image information of the object on the display unit, 10. The face captured image information of the object based on face arrangement guide information is acquired from the captured image information of the object stored in the predetermined storage unit. The character generation system of Claim 1.
The character generation system according to claim 9 or 10, wherein the predetermined face part whose arrangement is indicated by the face arrangement guide information includes at least a face outline.
A character generation method for generating a face of a character displayed on a display unit as being present in a virtual space as a caricature of an anime style based on face captured image information of an object using a computer,
The computer is
(A) obtaining the captured face image information of the object from the captured image information of the object stored in a predetermined storage unit;
(B) acquiring positional information of the feature points of the predetermined facial part from the face captured image information of the object acquired by the step (a) acquired by the step (a), and Obtaining color information of a predetermined pixel;
(C) generating a head shape model of the character based on the feature points of the predetermined facial part acquired in the step (a);
(D) setting similar skin color information similar to the skin color of the object to be the skin color of the head texture of the character based on the color information of the predetermined pixel acquired in the step (a); Selecting a similar anime-like part corresponding to the face part similar to the face part of the object based on the feature points acquired in (a), and setting the arrangement of the selected similar anime-like part Generating a facial part texture corresponding to the facial part by:
(E) The facial part texture generated in the step (b) is pasted on the head shape model of the character and is pasted on the head shape model generated in the step (c). Replacing the color information of the skin portion of the head texture with the similar skin color information set by the step (d);
A character generation method characterized by comprising:
The step (d)
(D1) Based on the position information of the feature points acquired in the step (b), the shape of the facial part of the object from among a plurality of animation-like parts prepared in advance corresponding to the facial part Selecting the similar anime-like parts similar to
(D2) selecting the similar skin color information similar to the skin color of the object from a plurality of skin color information prepared in advance based on the color information of the predetermined pixel acquired in the step (b); ,
(D3) Based on the position information of the feature points acquired in the step (b), the arrangement of the similar anime-like parts corresponding to the face part selected in the step (d1) is set and set Generating the facial part texture based on the arrangement of the similar anime-like parts;
The character generation method according to claim 12, further comprising:
The step (d1)
When the face part is composed of two left and right basic face parts, the position information of the feature point of either the left or right basic face part is converted into position information moved in line symmetry with respect to the face center line. The shape of the basic face part is estimated from the average value of the position information of the feature points of the two left and right basic face parts, and the similar anime-like parts corresponding to the estimated shape of the basic face part are determined as the plurality of animations. Select from the key parts,
When the face part is composed of one basic face part, the shape of the face part is estimated from position information of the feature points of the face part, and the similar animation tone corresponding to the estimated shape of the face part The character generation method according to claim 13, wherein a part is selected from the plurality of anime-like parts.
The step (d3)
In the facial part texture corresponding to the facial part, the position pasted on the head shape model is preset,
The nose and mouth similar animation-like parts and the mouth similar animation-like parts are arranged in positions in the face part texture of the nose and mouth,
The similar anime-like parts of the left and right eyes with respect to the face center line at the vertical position in the face part texture set in advance with the distance between the left and right eyes calculated from the position information of the feature points Is placed to be a line target,
The similar animated parts of the left and right eyebrows with respect to the face center line at the position in the vertical direction calculated from the position information of the feature point at the position in the left and right direction in the face part texture of the eyebrow set in advance. Arrange to be a line target,
15. The facial part texture corresponding to the nose, mouth, eyes, and eyebrows is generated based on the arrangement of the similar anime-like parts corresponding to the nose, mouth, eyes, and eyebrows. The character generation method described in 1.
The step (d3)
In the facial part texture corresponding to the facial part, the position pasted on the head shape model is preset,
Calculating the position of the similar anime-like part in the facial part texture corresponding to the facial part based on the positional information of the feature point acquired in the step (b);
15. The character generation method according to claim 13, wherein the facial part texture is generated based on the calculated position of the similar anime-like part.
The step (d3)
A similar model face part texture similar to the arrangement of the similar animation-like parts corresponding to the face part is selected from a plurality of model face part textures corresponding to the face part, and the selected similar model face part texture is The character generation method according to claim 15, wherein the character generation method is set as a facial part texture.
The step (c)
A basic head shape model similar to the face contour is selected from a plurality of contour shape models prepared in advance corresponding to the face contour, based on the feature points of the face contour,
A similar part shape model similar to the predetermined face part based on the feature points of the predetermined face part from among a plurality of part shape models prepared in advance corresponding to the predetermined face part other than the face contour Select
18. The character according to claim 12, wherein the head shape model is generated by combining the selected basic head shape model and the similar part shape model. Generation method.
The step (c)
13. The head shape model similar to a face contour is selected based on the feature points of the face contour from a plurality of reference shape models prepared in advance corresponding to the face contour. 18. The character generation method according to any one of items 1 to 17.
The computer is
(F) While displaying the face arrangement guide information indicating the arrangement of the predetermined face part and the object on the display unit, the object is imaged based on the face arrangement guide information, and the object The step of storing captured image information in the predetermined storage unit is further provided before (a),
In the step (a), based on the face arrangement guide information, the face captured image information of the object is acquired from the captured image information of the object stored in the predetermined storage unit. The character generation method according to claim 12, wherein the character generation method is characterized.
The computer is
(G) The step of inputting the captured image information of the object and storing it in the predetermined storage unit is further provided before (a),
In the step (a), the face arrangement guide information indicating the arrangement of the predetermined facial parts and the captured image information of the target object input in the step (g) are displayed on the display unit, 21. The captured face image information of the object is acquired from the captured image information of the object stored in a predetermined storage unit based on the face arrangement guide information. The character generation method according to any one of the preceding claims.
The character generation method according to claim 20 or 21, wherein the predetermined face part whose arrangement is indicated by the face arrangement guide information includes at least a face outline.
A program for causing a computer to execute processing for generating a character face displayed on a display unit as a presence in a virtual space as an animation-like caricature based on face captured image information of an object,
The program which makes the said computer perform the process which implement | achieves each means of the character generation system of any one of Claims 1 thru | or 11.